This invention relates generally to medical radiology and more particularly to methods and apparatus for radiographic imaging.
Radiographic imaging of all kinds, including computed tomography (CT) imaging, can detect small low contrast features. Thus, radiographic imaging has become important in medical practice, allowing medical practitioners to detect low contrast tumors and lesions in anatomical regions of soft tissue, including the brain and the liver. An important issue in radiology today concerns the reduction of radiation dose received by a patient during a CT examination without compromising image quality. Generally, higher radiation doses result in the ability to detect lower contrast smaller objects, while lower doses lead to increased image noise. Higher radiation doses also increase the risk of radiation-induced cancer. Thus, the ability to image low contrast objects at a low dose is desirable for diagnostic x-ray imaging methods.
The ability of a CT system to differentiate a low-contrast object from its background is measured by its low contrast detectability (LCD). LCD is measured using phantoms that contain low-contrast objects of various sizes. Phantoms that produce low contrast objects by using materials with different densities are useful for testing conventional energy integrating CT scanners. Phantoms that produce low contrast objects using energy sensitive materials allow performance testing for a dual energy scanner.
The low-contrast resolution of a CT scanner is generally defined as the diameter of an object that is just detectable at a given contrast level and dose. The contrast level is usually specified as a percentage of the linear attenuation coefficient of water. A sample specification with the current method might be “4 mm at 0.3% contrast for 10 mm slice thickness at 30 mGy CTDIvol dose.” Sometimes other dose metrics are used, such as the surface dose measured at the outer surface of the phantom or the Size Specific Dose Estimate {AAPM 2011}.
At least two LCD specifications are known. One known LCD specification is made at a single protocol using human observation. In this method, reconstructed images are viewed by one or more human observers to determine the smallest pin that, in the opinion of the observer, is visible. Another known LCD specification is made at a single protocol using a statistical method. In this method, an automated algorithm predicts the contrast required to detect a given size pin with a specified confidence interval from a flat “water” image.
These known LCD specifications characterize the performance of the CT scanner at only one protocol and one phantom size. Furthermore, the known LCD specifications do not characterize the performance of a CT scanner over an extended range. For example, only a portion of the full operating range of the scanner is characterized. It would therefore be desirable to provide methods and apparatus for characterizing the performance of a radiometric imaging apparatus such as a CT scanner at more than one protocol, over the full operating range of the imaging apparatus, or both.
Flux Index
At least some known commercial CT scanners operate over a wide range of protocols, each of which can have distinct contrast characteristics. The protocol parameters that affect contrast include scan time, tube current (mA), slice thickness, object diameter, tube voltage (kVp) and x-ray filter. Contrast is also significantly affected by non-linear reconstruction methods as well as the reconstruction pixel size and reconstruction filter. It is assumed herein that the tube voltage, the x-ray filter, the scan diameter and the reconstruction method, collectively comprising a core operating mode, are fixed and that the scanner, in that core operating mode, can be characterized by the CTDIvol dose index. Then the parameters (example values of which are given in parentheses) that directly affect the x-ray flux available for detection comprise scan time (0.25-2.0 sec/revolution), x-ray tube current (20-400 mA), slice thickness (0.5-10.0 mm), object diameter (20-50 cm), and dose index (CTDIvol)
At least one known LCD method uses a CTP515 low contrast module of the CATPHAN® phantom, available from Phantom Laboratory, Inc., Salem, N.Y. “Supra-slice” contrast sets are used but only the lowest 0.3% contrast set is typically reported.
There are at least two LCD measurement methods known to be used on commercial CT scanners. These methods are named the “human observer method” and the “statistical method.” We have compiled some recent reported measurements from the major CT manufacturers and collected them in Table 1. [NHS Purchasing and Supply Agency, Buyer's Guide, Computed Tomography Scanners, Reports CEP08007, CEP08027, CEP08028].
TABLE 1Recent Reported LCD Measurements from Major CT ManufacturersSource: NHS Purchasing and Supply Agency, Buyer's Guide, CT ScannersRef.Contrast SliceNum. on IndexScanner ContrastPin SizeDoseThicknessmAsFIG. 1Flux Index 500A0.3%4 mm 10 mGy10 mm9012900 400B0.3%5 mm 16 mGy10 mm1801414401000C0.3%2 mm 40 mGy10 mm350163600 400D0.3%5 mm7.3 mGy10 mm10518657These reported measurements show performance at only one point on the operating curve and that the operating point is different for each scanner, making performance comparisons invalid. This is shown as Prior Art FIG. 1 (based on it being based on previously reported measurements, not on it being presented an ExLCD graph) on an ExLCD graph 10 based on definitions of ExLCD Contrast Index and Flux Index described elsewhere herein.
Human Observer Method
In the Human Observer Method, LCD is determined by scanning a CATPHAN® phantom under selected protocol techniques and reconstructing the image or images of phantom. One or more human observers are then presented with the image or images of the phantom to render an opinion regarding the smallest object they believe is visible and therefore detectable for the 0.3% contrast set. For the reported measurements described above, it is not clear to the inventors whether a single observer or multiple observers were used. It is also not clear to the inventors how the specific protocol was selected to derive the reported specification.
Statistical Method
The statistical method for LCD avoids problems associated with human observers by relying only on noise measurements in a reconstruction. It does not use a phantom with actual contrast objects. Instead, it analyzes image noise in a specific manner that determines the amount of contrast needed to detect an object of a given diameter relative to the background with a stated level of confidence. Because the assessment is made by the computer and not a human observer, the method is repeatable and reproducible. However, the statistical method cannot differentiate contrast performance resulting from non-linear reconstruction methods since only a noise image is evaluated. The performance of the system relative to how well the original low contrast object is preserved thus cannot be determined, as is true of any noise analysis method that does not measure an actual object.
Quantum Noise Limited
An imaging system is said to be “quantum noise limited” if, for all practical purposes, the only source of image noise is the statistics of finite x-ray quanta and electronic noise is absent. Referring to graph 20 of prior art FIG. 2, the S/N (signal to noise) ratio is plotted as a function of relative x-ray Flux Index. In a log-log plot, the S/N ratio trace 22 for a quantum noise limited system is represented by a straight line having a slope of ½. If electronic noise (also known as “system noise”) is present, the overall S/N is significantly affected only for lower flux values as shown by trace 24 in FIG. 2.
With at least one LCD method known by the inventors to be in current use, a scanner is characterized with only one contrast measurement taken at a single protocol. This single measurement does not adequately characterize the contrast performance of the scanner. The single protocol measurement implies a contrast performance that follows a quantum noise limited curve defined by the single measurement. There is thus an inadequacy of the single protocol contrast performance curve. Additionally, this known LCD method does not adequately handle smaller pins that are affected by system blurring, i.e. the Modulation Transfer Function (MTF).
At least some known detectability methods that are based only on a noise analysis (such as the statistical method, noise power spectrum, simple-pixel standard deviation, and matched filter standard deviation) can overestimate the performance of a reconstruction process that alters the contrast of the test object. These known detectability methods use reconstruction processes that limit spatial bandwidth of both noise and object and do not account for changes in the assumed object. For example, assume that a small pin in an LCD test phantom is a cylinder with a 2 mm diameter and a contrast of 0.3%. If perfectly reconstructed, image pixels within the area of the pin have an average contrast of 0.3% and all pixels outside this region have an average contrast of 0%. However, the MTF of the system will blur the pin (especially at its edges) and spread some of its contrast into pixels beyond the original geometric boundary, resulting in a reduction in average contrast within the pin region.
Thus, it will be understood that inaccuracies of at least some known single protocol LCD methods result from human observer variation, finite pin size selections, selection of protocol, presence of system (electronic) noise; and/or system blurring (MTF) of smaller pins
The low contrast detectability (LCD) performance of a CT system is a critical performance characteristic, providing a measure of the ability of a scanner to produce high quality images at a low x-ray dose such as the lowest possible x-ray dose. Because the use of lower dose protocols in CT scanners is now of considerable importance, it is correspondingly desirable for LCD to be measurable over a wide range of protocols and body sizes. However, inaccuracies of the known prior art effectively prevent true differentiation of the contrast performance between CT scanners.
Automatic Exposure Control (AEC) systems for radiographic imaging systems such as CT are known to be in widespread clinical use. An objective of these systems is to reduce patient dose by allowing the CT system to determine and modulate an mA along a patient's Z axis as necessary to achieve a desired Clinical Image Quality (CIQ). A user determines or selects a CIQ necessary or desirable for the clinical application in terms of an Image Quality Metric (IQM) goal parameter provided by the CT vendor and the CT system is designed to produce the appropriate x-ray dose to achieve it. XY or angular modulation is also provided in at least some known CT systems, but AEC as used herein refers to Z axis modulation.
An important consideration for an AEC system is how the user specifies a desired CIQ. Depending upon the CT vendor, some known CT systems use a variety of IQMs. These methods include specifying a reference mA based on an nominal patient size chosen by the vendor, an image standard deviation, a noise index, or a reference image. However, methods known by the inventor to be in current use do not adequately describe CIQ, are not universal (i.e., the same values cannot be used on other make and model scanners), and may not track the desired CIQ with patient size. In addition, the use of different methods to determine an IQM increases confusion among technologists, increasing the likelihood of medical errors as well as making it more difficult to compare IQ and dose tradeoffs for different features and systems.
Size and contrast of an object, such as a lesion, that can be successfully identified with adequate sensitivity and specificity depend on many factors {Barrett 2004}. Object detectability is a significant component of clinical image quality and is related to dose applied and the image generation method used. It is well known that objects are more difficult to successfully identify as noise increases. Image noise is characterized as a mottle of pixel variations without any apparent consistent structure. CT image noise results from x-ray quanta as well as non-quantum sources. X-ray quantum noise is statistical photon noise that decreases inversely with the square root of the X-ray intensity, which in turn is proportional to the mA selection. Non-quantum noise includes electronic and electromagnetic sources and generally becomes a noticeable factor only at low x-ray flux levels with large patients. However, noise alone does not determine detectability, which is also influenced by how well an image generation system reproduces a scanned object within an image. The reproduction of the object is especially important when evaluating adaptive and model-based iterative image generation methods. Thus, an IQM based on detectability is better able to universally describe patient CIQ goals. The IQM goal metrics used by at least some known CT AEC systems are not universal.